Every year at its three-day conference in Salt Like City, Adobe presents a handful of not-quite-finished products to the 7,000 assembled marketers and engineers, to see what gets the best reaction. The “Sneaks” session gives a sense of what big capabilities and pain points Adobe wants to tackle in the coming year.
This year’s Sneaks, emceed by Wayne Brady, was almost entirely about analytics — not just making them more powerful, but making it easier for average marketers to interpret data that’s thrown at them.
As always, the audience was asked to vote for each Sneak by tweeting online. The clear winner was #SmartPic, an intelligent asset management feature that marketers and creatives can use to tag and track images. On its surface SmartPic is an auto tagging feature, which uses image analysis techniques developed initially for Photoshop to scan each uploaded image and tag it in context. Adobe’s team showed how their algorithms could figure out what season the photo was taken, what the setting was and what the subjects in it were doing – like biking or snowboarding.
But that was really only the start of the feature. SmartPic tracks how each image asset performs across the sites, apps and ads where it’s been used. That allows creatives to be able to tell which wintertime snowboarding image drove the most engagements and conversions over its lifetime. That got the marketers in the room on board.
“Another really cool thing about this is … it’s actually becoming metadata so I can use it for sorting and filtering. So I can come in here and find which assets are getting the highest number of impressions right now,” said Adobe research scientist Craig Mathis. “I can go from hundreds of thousands of assets down to just those few that are performing well with a particular segment.”
Another Sneak on display, DataTone, provided a Siri-like front end to Adobe’s analytics suite. If the user speaks or types a question about revenue, ROI, social interactions or whatever else, DataTone answers with a number – if it’s a simple question like “How many impressions did we get on the site yesterday?” – or with a detailed interactive chart if it’s a more complex question.
Most impressive was its ability to select the best visualization for the kind of data it was asked to analyze – with a heat map for revenue by region, or a line chart for revenue by time.
Other Sneaks focused on in-store analytics using mobile beacons, comparisons against industry benchmarks generated from Adobe’s large client database, and a pivot-table like interface where marketers could compare any analytics variable to any other for “freeform” data exploration. The latter was less popular, likely because it presented such an intimidating range of options.
Adobe’s presenters tried to highlight how fast each of the solutions worked. DataTone for example could crunch massive datasets and respond in seconds.
Not all the Sneaks will make it into the Marketing Cloud by next year, but they do give a sense of what Adobe’s priorities are – in this case making it easier for marketers to take the massive datasets they have and actually use them to improve strategy,
“We want to make sure that we give the most comprehensive datasets, but also insights,” said Carmen Sutter, product manager for Adobe Social, which focuses heavily on social media performance analytics. “It’s one thing to just throw numbers at somebody, but you need to be able to glean insights from that.”